The final form and organization of the CNS depends on interactions between cells; these interactions are dependent upon a number of specific classes of molecules, foremost among them being the cell recognition molecules (CRMs). In the mature CNS, CRMs continue to be crucial for interactions between cells as in, for example, synapse formation, neurite extension, and interactions between neurons and glia. We are exploring the possibility that abnormalities in CRM functions are involved in drug abuse and in neuropsychiatric disease. Two particular molecules, N-CAM and L1 antigen, are abnormal in the cerebrospinal fluid of schizophrenic patients. The differences do not represent an acute drug effect, but appear to be related to long-term drug treatment. Moreover, the abnormalities are present in a specific pattern, with the abnormality being seen in certain forms of each molecule: The 120 kD form of N-CAM is increased about two-fold in schizophrenic patients, while the 140 kD form of L1 antigen is decreased to a similar degree. Similarly, although smaller, changes in N-CAM are also seen in patients with bipolar mood disorder. These changes in CSF are reflected in abnormalities of expression of soluble N-CAM in hippocampal tissue from post-mortem samples from human patients. Recent data suggest that the abnormalities in CRMs seen in schizophrenia are related to long-term chronic drug treatment, i.e., on the order of at least several years, or to long durations of illness. Studies in progress have shown no effect of methamphetamine on expression of N-CAM, although chronic amphetamine alters expression of the synaptic protein synaptophysin. Abnormalities in TGFbs, which in some systems regulate expression of CRMs, do not appear to be responsible for these differences: TGFb2 was found to be increased in Parkinson's disease, but not in schizophrenia. Animal studies suggest that expression of N-CAM and L1 is altered in brain areas that are deafferented by injury, and that exogenous L1 can promote plasticity of dopaminergic neurites. These data suggest a possible role of CRMs in adaptation to chronic drug exposure and perhaps the predisposition to the development of neuropsychiatric disorders. In that the presence of neuropsychiatric disorder is a strong risk factor for the development of drug abuse, examination of abnormalities related to these disorders may provide clues related to vulnerability to the development of drug abuse.